Hacker News Reader

Comment by codethief | Hacker News Reader

> But it's an axiom.

I don't agree. If light traveled instantaneously in one direction, then if we looked in the opposite direction (where such light originates from) we would be seeing stars and galaxies at much more recent time (now). Also, their light would have traveled a much longer distance (due to the ongoing expansion of space) and so would be redshifted much more. All in all what we see on the sky would look very differently in terms of redshifts, matter distribution and so on – unless of course there's no isotropy, i.e. unless there's some cosmos-sized conspiration that fine-tuned matter distribution, distances etc. in such a way that the universe looked isotropic to us even though it is not.

View on HN

Good point! Simplified (trying to explain to myself), if light travelled slower in one direction than others, some objects would be less far away from us and each other and thus their gravitational effects would be different.

Basically, our cosmological horizon would be significantly closer in one direction than others - and all objects (= mass) in that direction would be as well.

There might be a very specific way in which matter could be distributed so that this is actually true but indistinguishable from uniform light speed in all directions, from our specific point of view. But I'd expect the effects to be noticeably different even within a fairly limited time range (decades, perhaps centuries).

You can change other things to get the same result.

You can build a consistent theory of aether by asserting that everything (including you) doppler shifts just so and there is a 0 speed through the aether. You get a complicated theory that is not experimentally distinguishable and is incredibly difficult to use and noone would have any idea how to extend to QM or gravity because it's so cumbersome, but it's consistent and mechanics works.

You could easily reframe the frame of reference in such a theory where the lab is moving at 0.999c absolute (which exists in this theory despite being immeasurable) as light moving very slowly in one direction and near instantly in the other.

> if we looked in the opposite direction (where such light originates from) we would be seeing stars and galaxies at much more recent time (now)

If the preferred/non-preferred directions were "toward you" and "away from you", then rotating yourself to look in the "opposite direction" wouldn't make a difference.

(Note: not a physicist)

> Also, their light would have traveled a much longer distance (due to the ongoing expansion of space) and so would be redshifted much more.

I need to correct myself: If light traveled instantaneously, the expansion of space wouldn't matter because in the very instant in which light would be traveling space wouldn't be expanding. -> No redshift at all.

I get what you are saying, it suggests there is no preferred direction, but it is just an inference. The point is one cannot create an experiment to measure the speed of light other than a round trip average speed.

Wasn't that the whole idea behind Michelson-Morley ( https://en.wikipedia.org/wiki/Michelson%E2%80%93Morley_exper...)?

Yes, that is an example of a measurement of the two-way speed of light. What is impossible, however, is a measurement of the one-way speed of light.

M-M compared the speed in one direction to the speed at right angles. That's not quite the same thing as a one-way measurement, but it's also not the same thing as a two-way measurement along the same path.

Right, but the way that they measured the speed in any direction was by making a two-way measurement. They sent a pulse of light out to be reflected and timed how long it took for it to return.

I imagine that a ring-laser gyro would return some pretty strange results in a universe where c was anisotropic.

Nope. A ring–laser gyro sends two light waves in opposite directions around a loop. Some of the light from both beams is let out at one point on the perimeter and the interference between them is measured. It can be any shape you like, as long as the beams trace out the perimeter of some area. When the gyro is rotating, one of the beams has to travel around the perimeter plus some extra distance due to the rotation, while the other beam has to travel the perimeter less that same distance. This causes the interference pattern to shift. The amount of the shift is proportional to the area inside the perimeter, the amount of rotation, the speed of light, and the frequency of the light that you are using.

When the speed of light is anisotropic, we have to replace the constant c with some horrible integral which makes the math a lot worse. However, it is important to recognize that the overall time to go around the perimeter clockwise and counterclockwise is the same: both beams go the same distance to the left as they go to the right. As long as the leftwards speed of light and the rightward speed of light average out to the usual value, then the result will work out to be the same as when the speed of light is a constant with the usual value in all directions.

Any time you send a pulse of light out and back, the time to travel the distance d will be αcd+βcd=2cd, where α and β show the relationship between the speeds of light in those two directions. We generally assume that α=β=1.0, but as long as α+β=2.0, then everything we can measure will work out to be exactly the same.

I don't remember that experiment measuring the speed, but merely showing (via lack of a beat pattern) that there was no difference in the speed, whatever it was, in two paths at right angles to each other.

This does not imply that the speed of light could depend on direction, though. If that were the case, we could see it by rotating a simple Michelson interferometer.

Red shift would be different in various directions then. We would also see fewer young galaxies / stars in the direction in which light flies faster.

That would imply an axiom that forces/light/whatever are on a gradient we operate against and does not change based on reference point.

I think the heart of the what the GP is getting at is no matter how you define it, you can measure things smaller than your measuring equipment can detect.

Reply on news.ycombinator.com